Background
Multicenter (9 hospitals, China), open-label, randomized phase 3 superiority trial (DRAGON-01) in treatment-naive adults with gastric adenocarcinoma and laparoscopically confirmed peritoneal metastasis without extraperitoneal disease (238 randomized 2:1; 222 treated). Tested whether adding intraperitoneal (IP) paclitaxel — which achieves high intracavitary drug concentrations — to intravenous paclitaxel plus S-1 improves overall survival.
Results
Interventions and follow up: Arm A (IP): paclitaxel 50 mg/m² IV plus 20 mg/m² intraperitoneally on days 1 and 8 of each 21-day cycle, with oral S-1 80 mg/m²/day on days 1–14
Arm B (PS): paclitaxel 70 mg/m² IV on days 1 and 8, with oral S-1 80 mg/m²/day on days 1–14
Primary endpoint: overall survival
mFollow up: 72.2mo
Results: OS: 19.4mo vs 13.9mo (IP vs PS); HR 0.67, 95%CI 0.50-0.90, P=.01
PFS: 11.2mo vs 7.2mo; HR 0.72, 95%CI 0.54-0.96
Arm B (PS): paclitaxel 70 mg/m² IV on days 1 and 8, with oral S-1 80 mg/m²/day on days 1–14
Primary endpoint: overall survival
mFollow up: 72.2mo
Results: OS: 19.4mo vs 13.9mo (IP vs PS); HR 0.67, 95%CI 0.50-0.90, P=.01
PFS: 11.2mo vs 7.2mo; HR 0.72, 95%CI 0.54-0.96
Adverse events
Grade 3-4 (any): 38.5% vs 41.9% (IP vs PS)
Treatment-related deaths: none in either arm
Treatment-related deaths: none in either arm
Conclusions
Adding intraperitoneal paclitaxel to intravenous paclitaxel plus S-1 significantly prolonged overall survival (median 19.4 vs 13.9mo) and PFS in treatment-naive gastric cancer with peritoneal metastasis, without increasing severe toxicity. DRAGON-01 provides the first phase 3 evidence of an OS benefit for IP paclitaxel in this poor-prognosis population.
Key Limitations
Open-label design; conducted entirely in China across 9 sites, limiting generalizability; modest sample (238 randomized) with a smaller 2:1 control arm; IP delivery requires intraperitoneal port placement, adding procedural burden; ChiCTR-registered with no concurrent FDA/EMA registration pathway.
Clinical Context
Gastric cancer with peritoneal metastasis carries a dismal prognosis, and IV paclitaxel plus S-1 is a standard option. The prior Japanese PHOENIX-GC phase 3 of IP paclitaxel missed its primary endpoint on the primary analysis (positive only on adjusted sensitivity analysis). DRAGON-01 is the first phase 3 to show a statistically significant OS benefit for IP paclitaxel. There is no FDA/EMA approval for IP paclitaxel in this setting, and NCCN/ESMO gastric guidelines do not currently recommend intraperitoneal chemotherapy outside clinical trials.